The line works in J6.02 and produces  0 2 24$'a' .
Whatever the bug was was apparently fixed in J6.02.



----- Original Message -----
From: Henry Rich <[EMAIL PROTECTED]>
Date: Saturday, August 11, 2007 10:42
Subject: RE: [Jprogramming] Crash in J6.01, related to monad/dyad and fill
To: 'Programming forum' <[email protected]>

> I have a verb, rather long I'm afraid, but it is only one line:
> 
> 
> efr =: 44100&efr : (('fmt ' , (16 0 0 0{a.) , ([: (1&(3!:4)) 1 , 
> ]) , 2&(3!:4)@[ , ([:
> (2&(3!:4)) 2 * *) , [: (1&(3!:4)) 16 ,~ 2 * ]) 2:)
> 
> When I make that definition, the following line crashes J 
> (vanishes without a trace):
> 
> (0$a:) (44100&efr 'xxx'"_)"0 1 (0 2$0)
> 
> 
> I have tried to cut down the verb more, but with little success.
> I have made some interesting observations, though:
> 
> 0.  If I change the definition of efr by removing the 
> initial 
>   44100&efr :
> bit, it doesn't crash.
> 
> 1.  If I execute
> 
>  a: (44100&efr 'xxx'"_) 0 0
> 
> it doesn't crash.  Interesting because I think this is what 
> should really be
> executing in the failing case.
> 
> 
>    9!:14''
> j601/2006-11-17/17:05
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to